Our Focal Areas

Data Science in healthcare leverages advanced analytics and AI to extract meaningful insights from vast amounts of medical data. At the AI Center for Health, we integrate diverse data sources, including clinical records, imaging-related information, and omics data (e.g. genomics, transcriptomics, proteomics), to drive cutting-edge research and innovation. Our expertise lies in developing machine learning and deep learning models for disease classification, patient stratification, outcome prediction, and survival analysis. 

A key focus is the integration of multimodal data, combining heterogeneous information to improve diagnostic accuracy and personalize treatment strategies (precision medicine). In collaboration with the IT department, we support building and deploying data infrastructures that integrate diverse sources—including clinical records, radiology, treatment data, and administrative information—ensuring a comprehensive and accessible ecosystem for research and innovation. We develop and implement tools for safe data handling, including anonymization techniques, always adhering to GDPR regulations. Patient privacy is our highest priority, and our infrastructure is designed to uphold the highest standards of security and ethical AI use in healthcare.

Computer vision enables AI systems to interpret and analyze medical images and videos, playing a crucial role in modern healthcare. At the AI Center for Health, we work with diverse imaging modalities, including radiology (MRI, CT, X-ray), histopathology, microscopy, and endoscopy videos, to develop AI-powered tools for clinical decision support.

Our models tackle key tasks such as image classification, segmentation, object detection, anomaly detection, and feature extraction (e.g., radiomics) to enhance disease diagnosis, outcome prediction, and toxicity assessment. A growing focus is on foundation models and visual-language models, which promise more generalizable, robust, and safer AI solutions for healthcare. These advanced AI approaches aim to reduce biases, improve interpretability, and ensure reliability across different patient populations and imaging settings, ultimately supporting clinicians in making more accurate and efficient diagnoses.

Natural Language Processing (NLP) enables AI to understand, process, and extract valuable insights from medical text, making it a critical tool in healthcare. At the AI Center for Health, we leverage NLP to analyze vast amounts of unstructured data from Electronic Health Records (EHRs), clinical notes, radiology reports, and scientific literature. This helps transform hidden textual information into structured data, improving disease diagnosis, patient stratification, and personalized treatment plans. Recent advancements in Large Language Models (LLMs) are particularly promising, allowing for more accurate information retrieval, summarization, and integration of textual data into precision medicine models.

In addition, AI is not only revolutionizing clinical care but also reshaping hospital operations to enhance efficiency and the overall patient experience. Our focus extends also to the development of innovative AI-driven solutions to optimize hospital workflows, improve resource management, and support both clinical and non-clinical staff. This includes the development of virtual assistants for enhanced customer care, automated reporting tools to facilitate clinician-patient interactions, and AI-powered systems to reduce administrative burdens. 

By harnessing these state-of-the-art AI techniques, we aim to enhance decision support systems, streamline clinical workflows, and unlock new possibilities for data-driven, patient-centered care.

Synthetic data, created using AI models, mimics real-world patient data while preserving privacy, enabling secure data sharing for open research. At the AI Center for Health, we use generative AI technologies to generate realistic datasets for training AI models, creating synthetic control arms for clinical trials, and developing digital twins to simulate patient responses for precision medicine. These innovations make clinical trials smarter, faster, and more cost-effective while ensuring robust decision support systems. While we showed the significant impact of GenAI and synthetic data in hematology and oncology, their applications are rapidly expanding to other medical fields, offering new opportunities for AI-driven healthcare advancements.